Analysis And Classification Of Human Arm Skin Texture Feature

With the improvement of living standard,the problem of skin texture has aroused more and more people's concern.And the skin aging problem also attracts the attention of the general medical personnel.Quantitative assessment of skin aging has become an important problem skin medical region.At present,the evaluation criterion is still depended on the doctor's subjective judgments.And the emergence of some modern optical technology evaluation methods has also presented,such as multiphoton microscopy,spectroscopy,etc.But these optical methods would damage skin,and they are hard to evaluate the skin directly.In this thesis,the image processing combined with traditional optical imaging were used to study the surface texture image feature in different ages skin.The quantitative analysis of the aging skin were uncover and the skin texture image classification of three groups of young,middle-aged and elderly were implementThe Tamura texture feature combined with wavelet transform were proposed for quantitative evaluation of skin aging.The parameters,such as coarseness,contrast and variance were extracted by this method.Meanwhile,Fourier transform were used to extract the orientation index of the skin image.All the four parameters coarseness,contrast,variance and orientation index were used for assessment the skin aging process,and then classified the skin texture image based on support vector machine(SVM).In our study,texture features were used to evaluate the skin aging,especially,the orientation index.The results indicated that the texture analysis can well quantitative evaluate skin aging,and obtain the better classification accuracy in the three groups of young,middle-aged and elderly groups.